Strong positive density-dependence should lead to a loss of diversity, but warning-colour and Müllerian mimicry systems show extraordinary levels of diversity. Here, we propose an analytical model to explore the dynamics of two forms of a Müllerian mimic in a heterogeneous environment with two alternative model species. Two connected populations of a dimorphic, chemically defended mimic are allowed to evolve and disperse. The proportions of the respective model species vary spatially. We use a nonlinear approximation of Müller's number-dependent equations to model a situation where the mortality for either form of the mimic decreases hyberbolically when its local density increases. A first non-spatial analysis confirms that the positive density-dependence makes coexistence of mimetic forms unstable in a single isolated patch, but shows that mimicry of the rarer model can be stable once established. The two-patch analysis shows that when models have different abundance in different places, local mimetic diversity in the mimic is maintained only if spatial heterogeneity is strong, or, more interestingly, if the mimic is not too strongly distasteful. Therefore, mildly toxic species can become polymorphic in a wider range of ecological settings. Spatial dynamics thus reveal a region of Müllerian polymorphism separating classical Batesian polymorphism and Müllerian monomorphism along the mimic's palatability spectrum. Such polymorphism-palatability relationship in a spatial environment provides a parsimonious hypothesis accounting for the observed Müllerian polymorphism that does not require quasi-Batesian dynamics. While the stability of coexistence depends on all factors, only the migration rate and strength of selection appear to affect the level of diversity at the polymorphic equilibrium. Local adaptation is predicted in most polymorphic cases. These results are in very good accordance with recent empirical findings on the polymorphic butterflies Heliconius numata and H. cydno.
All Science Journal Classification (ASJC) codes
- Statistics and Probability
- Modelling and Simulation
- Immunology and Microbiology(all)
- Biochemistry, Genetics and Molecular Biology(all)
- Agricultural and Biological Sciences(all)
- Applied Mathematics